Los Alamos National Laboratory researchers have, for the first time, successfully used artificial intelligence to decode data from astrophysical events – specifically, neutron star mergers and X-ray emissions – to infer the forces governing interactions within atomic nuclei. The breakthrough, detailed in a recent publication in Nature Communications, connects observations of cosmic phenomena to the fundamental physics of matter at the quantum level.
The research team, collaborating with scientists at the Technical University of Darmstadt in Germany, applied machine learning algorithms to data collected from the 2017 detection of gravitational waves produced by a binary neutron star merger, as well as data from telescopes studying neutron star X-ray emissions. This allowed them to constrain nuclear couplings, which describe the strength of the forces between neutrons and protons.
“This research represents the first time in the field that we’ve been able to robustly connect the macroscopic and microscopic realms and infer the interactions among neutrons and protons directly from astrophysical data,” said Ingo Tews, a physicist at Los Alamos. “Using artificial intelligence and machine learning, our framework made it possible to take data from remarkable astrophysical phenomena and infer the complicated physics of nuclear forces.”
Neutron stars, formed from the collapsed cores of massive stars, represent an extreme environment for studying nuclear physics. Their immense density – exceeded only by black holes – creates conditions where protons can convert into neutrons, and free neutrons can escape, potentially forming heavy elements through a process known as the rapid neutron-capture process, or “r process.” Understanding this process is crucial to explaining the origin of heavy elements like uranium and plutonium.
The computational challenges of modeling these interactions are significant. According to the Los Alamos team, applying numerous models of neutron interactions to incredibly dense neutron stars would be “computationally intractable,” with solutions requiring extensive processing time even on thousands of CPU cores. The AI-driven approach offers a pathway to overcome these limitations.
Isak Svensson, a scientist at the Technical University of Darmstadt and co-lead author of the study, explained that their framework “opens a new window into the strong-force physics of neutrons and protons and its effects on neutron stars,” enabling a direct link between neutron star observations and the behavior of dense matter.
Related research at Los Alamos focuses on the conditions within collapsing stars and the resulting gamma-ray bursts. Scientists there propose that high-energy photons produced during these events can dissolve the outer layers of stars into neutrons, creating the necessary conditions for heavy element formation. Free neutrons, still, have a short lifespan of approximately 15 minutes, making the timing and abundance of their production critical.
The work builds on decades of research into nuclear and particle physics, astrophysics, and cosmology at Los Alamos, encompassing studies ranging from the fundamental constituents of protons and neutrons to the behavior of matter in the immediate aftermath of the Massive Bang. The laboratory’s theoretical division continues to investigate the equations of state and structure of massive neutron stars, as well as the gravitational waves emitted during their mergers.
The team’s findings come as scientists continue to investigate X-ray superbursts, rare and powerful explosions observed on some neutron stars. These events, 1000 times more energetic than typical X-ray bursts, remain a source of ongoing study for researchers at Los Alamos and elsewhere.
